Fairness Meets Privacy: Integrating Differential Privacy and Demographic Parity in Multi-class Classification

TL;DR

This paper demonstrates that differential privacy can be effectively integrated with fairness constraints in multi-class classification, achieving strong privacy and fairness guarantees with minimal trade-offs.

Contribution

It introduces the DP2DP postprocessing algorithm that enforces demographic parity and differential privacy simultaneously, challenging the notion that privacy compromises fairness.

Findings

DP2DP converges to demographic parity at near-optimal rates

The algorithm achieves state-of-the-art accuracy, fairness, and privacy trade-offs

Experimental results validate theoretical guarantees on synthetic and real datasets

Abstract

The increasing use of machine learning in sensitive applications demands algorithms that simultaneously preserve data privacy and ensure fairness across potentially sensitive sub-populations. While privacy and fairness have each been extensively studied, their joint treatment remains poorly understood. Existing research often frames them as conflicting objectives, with multiple studies suggesting that strong privacy notions such as differential privacy inevitably compromise fairness. In this work, we challenge that perspective by showing that differential privacy can be integrated into a fairness-enhancing pipeline with minimal impact on fairness guarantees. We design a postprocessing algorithm, called DP2DP, that enforces both demographic parity and differential privacy. Our analysis reveals that our algorithm converges towards its demographic parity objective at essentially the same rate (up logarithmic factor) as the best non-private methods from the literature. Experiments on both synthetic and real datasets confirm our theoretical results, showing that the proposed algorithm achieves state-of-the-art accuracy/fairness/privacy trade-offs.